The honest practitioner's take on agentic AI on Kubernetes | BRK222

Author: Microsoft Developer - Bewertung: 1x - Views:56

Agentic AI workloads don't behave like normal services. They're stateful, bursty, multi-step, and often span more than a single cluster. Most teams figure this out the hard way. This session is for developers building agentic AI in production. We'll show you how to get the most out of Kubernetes for training, inference, and agent orchestration, covering purpose-built tooling, managed options, open-source inference at scale, and AI-assisted dev tools that support real-world AI operations.



Seating for this session is first-come, first-served. Add it to your schedule to plan your day and arrive early to secure a spot.



To learn more, please check out these resources:

* https://aka.ms/build26/BRK222





𝗦𝗽𝗲𝗮𝗸𝗲𝗿𝘀:

* Lachlan Evenson





𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻:

This is one of many sessions from the Microsoft Build 2026 event. View even more sessions on-demand and learn about Microsoft Build at https://build.microsoft.com



BRK222 | English (US) | Cloud platform & data





Breakout | (300) Advanced





#MSBuild



Chapters:

0:00 - Agenda Overview and Enthusiasm for Agentic Systems

00:09:41 - AI Workload Types: Inference, Training, and Agentic Systems Explained

00:12:36 - Challenges of scheduling massive AI workloads and need for gang scheduling

00:14:12 - Transition to layering AI platforms on top of Kubernetes

00:25:29 - Architecture Overview: Integration of AnyScale Runtime with Azure

00:30:06 - Launching and Managing AnyScale Workspaces and Clusters

00:35:05 - Conclusion: Agentic Orchestration and Future of Autonomous Agents

00:38:19 - Launching demo of Azure KARS: secure sandbox using workload identity

00:47:36 - Final segment on AKS Claw and building agentic workloads atop Kubernetes